Episode Summary:
In this episode of BetterTech on How Shadow IT Breaks AI Transformation, host Colin McCarthy sits down with JD Fetterly, Senior AI Strategy and Value Realization Manager at Within, to explore why the biggest bottleneck in AI adoption often has little to do with the model itself. Drawing on two decades of experience across GE, Google, Apple, and Within, JD explains how process transformation has been the common thread throughout his career and why organizations need to understand how work actually happens before deciding where AI should be deployed.
The conversation moves from data governance myths, to practical agentic use cases like automated meeting prep and client briefs, to the human-agent operating model Within uses to decide what gets automated and what stays human. JD and Colin also dig into “shadow IT” and hidden workflows that leadership rarely sees, what separates failed transformation programs from successful ones, and why the best tech leaders are the ones who give their best work away to grow their teams.
Show Notes:
- Why “our data isn’t ready” is usually a description problem, not a data problem, and a real-world example of an agent-built customer meeting brief that cut prep time in half even while imperfect.
- How Within’s human-agent operating model and skills registry decide what agents own versus what humans keep, and how this changes the way teams work.
- Why leadership rarely sees the “hidden factories” like manual workarounds, unofficial spreadsheets, and How Shadow IT Breaks AI Transformation.
- Why AI transformation programs fail for the same reasons legacy transformations did: mapping too slowly against a business that is changing in real time.
- JD’s take on which business function AI disrupts next, why it’s less about one function and more about knowing where to point agents, and what separates good tech leaders from great ones: giving away the work you love so your team can grow into it.
This episode is hosted by Colin McCarthy, Host, BetterTech
